Industrial Fault Localization Using Variable Perturbation Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial manufacturing plants face challenges in accurately identifying the root cause of faults due to nonlinear interactions among process variables and nonstationary behavior, leading to inefficiencies and unplanned downtime.
Innovation Solution
A method and system utilizing multivariate time-series data, soft-sensors, and a multi-level variable perturbation approach to identify dominant variables contributing to faults, employing hardware processors for data preprocessing, fault analysis, and localization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional fault detection techniques are used, then fault detection capability is provided, but fault localization precision deteriorates due to nonlinear interactions among process variables
Solution Approach 1:
The patent segments the complex fault localization problem into multiple hierarchical levels: (1) fault detection level using residual analysis, (2) fault isolation level using directed graphs and cut-sets, and (3) root cause identification level using variable perturbation. This segmentation allows each level to handle specific aspects of the problem independently, reducing overall modeling complexity while maintaining localization precision.
Solution Approach 2:
The patent introduces soft sensors as intermediary components that bridge the gap between raw process measurements and fault analysis. These soft sensors compute unmeasured or difficult-to-measure variables based on available measurements, thereby simplifying the fault localization process without sacrificing precision by acting as mediators between the complex process system and the diagnostic algorithms.
2Reliability
If comprehensive monitoring of all components is implemented, then fault detection capability improves, but system complexity and computational burden increase
Solution Approach 1:
The patent extracts and focuses only on the critical variables and components that significantly contribute to fault occurrence using the directed graph model and cut-set analysis. Instead of comprehensively monitoring all components equally, the system identifies and extracts the minimal set of critical variables that, when monitored, provide sufficient fault detection capability while reducing system complexity.
Solution Approach 2:
The patent applies local quality by assigning different monitoring intensities and analytical methods to different parts of the system based on their fault criticality. Critical components identified through cut-set analysis receive enhanced monitoring and more sophisticated diagnostic algorithms, while less critical components use simpler monitoring approaches, thereby optimizing the balance between reliability and complexity.
3Measurement precision
If detailed analysis of all process variables is performed, then root cause identification accuracy improves, but computational time and resources increase
Solution Approach 1:
The patent performs preliminary action by pre-computing the directed graph model, identifying cut-sets, and determining critical variables before actual fault occurrence. This preliminary structuring of the analysis framework allows the system to quickly localize faults and identify root causes when faults occur, without performing exhaustive analysis of all variables in real-time, thereby reducing computational time while maintaining accuracy.
Solution Approach 2:
The patent applies partial action by performing detailed analysis only on the subset of variables identified as critical through the directed graph and cut-set analysis, rather than analyzing all process variables equally. This selective approach concentrates computational resources on the most relevant variables, achieving high root cause identification accuracy with reduced computational time and resources.
Data Source
AI summary
Existing systems for fault detection and classification have the disadvantage that they have limited or no capability for fault localization and root cause identification, probably due to the challenges associated with modeling the nonlinear interactions among process variables and capturing the nonstationary behavior that is typical of most industrial processes. The disclosure herein generally relates to industrial manufacturing systems, and, more particularly, to method and system for localization of faults in an industrial manufacturing plant. The system uses a perturbation based approach for fault localization, in which the system determines variables having dominant effect on identified faults, in terms of a perturbation score calculated for each of the variables.